
This is a port of my SD1.5 embedding to XL LoRa. Comic book drawing style with bold, straight lines, washed colors and grim character.
Trained on 45 images, in 4000 steps. Which took about 2.5 hours on 3090. Input images were generated in SD1.5, upscaled and heavily edited.
SDXL version was trained on base model and Pony on AutismMix. It works fairly well with merges, but I would avoid photography focused fine-tunes. Especially for Pony. Quality/score tags are unnecessary. Use 0.5 to 1.0 weights depending on checkpoint and desired effect.
Have fun.
描述:
训练词语: s_darkcore style
名称: darkcore_pny.safetensors
大小 (KB): 96410
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success